Abstract: Active Shape Models（ASM） is one of powerful tools for face alignment，face recognition and facial expression recognition.However，the performance of ASM is often influenced by some factors such as the initial location，illumination and so on.This paper proposes an improved active shape models.First，the position of eyes which provide roughly the initialization position for point distribution model of ASM is detected in face region using Point Contour Detection Method（PCDM）.Second，a new method constructed local appearance model is developed.This improved method is evaluated on JAFFE face database and the experimental results show that the improved ASM improves the accuracy of face features location.Finally，an artificial neural network is constructed to recognize facial expression and gets a good recognition rate.